Introduction to Machine
Vision for New Users
Part 1 of a 3-part webinar series: Introduction to Machine Vision
Track, Trace & Control Solutions
© 2010 Microscan Systems, Inc.
About your Instructors
Dr. Jonathan Ludlow
Machine Vision Product Manager
Juan Worle
Technical Training Coordinator
© 2010 Microscan Systems, Inc.
Today’s Objectives
By the end of this Webinar, you will know:
 What is Machine Vision
 How Machine Vision can be used in
different applications
 What makes a Machine Vision system
© 2010 Microscan Systems, Inc.
Today’s Topics
Today we will discuss:
 What is Machine Vision?
 Why Use Machine Vision?
 Parts of a Machine Vision system
© 2010 Microscan Systems, Inc.
What is Machine Vision?
Automatic extraction of information from digital images.
The most popular Machine Vision applications include:
1.
2.
3.
4.
Measure: measure a part and check to tolerances.
Decode: 1D codes, 2D symbols, and OCR reading.
Count: find the number of parts or features on a part.
Locate: report the position and orientation of a part.
It’s not magic!
Arthur C. Clarke’s Third Law:
Any sufficiently advanced technology is
indistinguishable from magic.
© 2010 Microscan Systems, Inc.
1. Measure
Automated measurement, and then check to specified tolerances.
Measuring a spark plug for proper gap tolerance
© 2010 Microscan Systems, Inc.
2. Decode
Decode 1D or 2D Symbologies, read OCR.
 Record information for historical data
 Use information for immediate action
 Validate the data for correctness
OCR on cans used
to identify contents
© 2010 Microscan Systems, Inc.
1D and 2D symbols
DPM Data Matrix
3. Count
Find a number of parts or features on a part.
 Missing parts/proper assembly
 Presence: is it there or not?
 Identify quantity
Count the number of
sodas in an 8-pack
Counting the holes
drilled in a machined
block of aluminum
© 2010 Microscan Systems, Inc.
4. Locate
Report the position and orientation of a part.
 Locate position and orientation then check to specified tolerances
 Report information to another device for guidance
 Use for alignment of other Machine Vision tools
 Find a unique pattern to identify the part
Search for a pattern to
identify a product
© 2010 Microscan Systems, Inc.
Locate tools can align
images for other tools
Why Use Machine Vision?
 Reduce Defects
 Track WIP
 Increase Yield
 Comply with Regulations
 Machines don’t get tired
© 2010 Microscan Systems, Inc.
Parts of a Machine Vision System
Example component
system
Part
The object to be measured or inspected
Examples of integrated
systems
1. Lighting
The most critical piece, illuminates the part
2. Lens
Delivers image to the sensor
3. Sensor
Captures light and converts it to a digital image
4. Vision Processing
Looks at the picture to extract features
5. Communicate
Report the results to the world, can be data or discrete I/O
2D imagers use the same parts
as a Machine Vision system
© 2010 Microscan Systems, Inc.
1. Lighting and the part
Proper lighting is essential to a successful Machine Vision application.


Why is lighting so important?
Select the right light and geometry:
˗ Reveal the part’s features
˗ Minimize everything else
If the feature cannot be seen,
it cannot be measured
© 2010 Microscan Systems, Inc.
2. Lens
Delivers the image to the sensor.
 Determines the Field of View, Depth of Focus, and Focal Point
 Interchangeable: C-mount
 Fixed: auto-focus
Depth of Focus
12mm: larger Field of View
Lens &
extension tubes
25mm: magnified image
© 2010 Microscan Systems, Inc.
3. Sensor
Captures light and converts it to a digital image.
Sensor is
inside the
camera
.3MP sensor
2MP sensor
Sensor pixels collect light
© 2010 Microscan Systems, Inc.
4. Vision Processing
Extracting useful information from images
 Acquire image: Collect the image that was captured by the sensor/lens
 Image Pre-processing: Modify the image to make features stand out
 Image Analysis: Extract features from the image
 Geometry & Tolerance: Measure features and compare to specification
 Results: Communicate Pass/Fail
Counting PCB pads
using blobs
© 2010 Microscan Systems, Inc.
Measure using
edge tools
Locate a part by searching
for unique features
5. Communicate
Report the inspection results.
 Discrete I/O: Trigger inspection, drive indicator lights, PLCs, diverters
 Data: send data over RS-232, Ethernet, Ethernet/IP…
A PLC can collect data
and discrete I/O signals
Integrated systems have
I/O built in
Electrical signals can be sent to a
diverter to remove faulty items, and
to a stack light to warn operators
HMI screens can display Machine
Vision inspection status
© 2010 Microscan Systems, Inc.
Parts of a Machine Vision System
Putting it together: Check fill level and cap installation
1.
2.
3.
4.
5.
6.
Trigger the inspection (Communicate)
Illuminate the bottle (Lighting)
Capture the image (Lens and Sensor)
Run Machine Vision tools (Vision Processing)
Show results to an HMI (Communicate)
Send a signal to a reject device (Communicate)
3
5
2
4
1
© 2010 Microscan Systems, Inc.
6
Introduction to Machine Vision for New Users
Conclusion
 Machine Vision can be used to automate manual tasks
– Measure
– Decode
– Count
– Locate
 There are five key components to a vision system:
– Part/Lighting: the features to be extracted must stand out consistently
– Lens: select the proper lens combination to deliver best Field of View and Depth of
Focus to the sensor
– Sensor: the sensor type and number of pixels is important to acquire a good image
– Vision Processing: extract useful information from an image
– Communicate: send the results in a way that is useful
© 2010 Microscan Systems, Inc.
© 2010 Microscan Systems, Inc.
Next session….
Choosing the Right Machine Vision Applications
 Successful Machine Vision Applications
 Questionable Machine Vision Applications
 Machine Vision Hardware Platforms
Introduction to Machine Vision Webinar series:
Today: Introduction to Machine Vision for New Users
December 10: Choosing the Right Machine Vision Applications
December 17: Machine Vision Tools for Solving Auto ID Applications
© 2010 Microscan Systems, Inc.
© 2010 Microscan Systems, Inc.